MYTHBUSTING Takes on Shielded Cabling Herb Congdon Brian Davis
Objectives Take on myths of shielded cabling and use modern-day science to show what's real and what's fiction - through trial and error actually demonstrate it.
Background Shielded is century-old technology (1880) Dating back to early coaxial cabling Shielded Twisted-pair Cabling is over 25 years old Dating back to IBM Type I Cabling
Background Wide industry experience with IBM Type 1 That experience was, admittedly, mostly negative Partly why the industry preference moved to unshielded in North America ( cheap and easy over high performance )
Background Shielded technology has innovated, evolved and developed Evidence of that has not been widely seen in North America due to the limited use of shielded cabling
Background So, the previous convictions of IBM Type 1 are still applied to all shielded cabling regarding: Cable Termination The Antenna Effect Ground Loops The Baked Potato Effect
Cable Termination
Myth: Cable Termination Terminating a shielded cable with a shielded connector takes twice as long as terminating an unshielded cable with an unshielded connector.
IBM Type 1 Cable Research Two shielded twisted pairs not four pair 22 AWG solid copper conductors Braided shield IBM Data Connector Hermaphroditic 4 pole Large
Research Data Connector Termination Procedure: 13 steps! snip & cut, slide, twist, cut & trim, clamp, slide, insert (x4), pull, align & squeeze, ground, bend, slide, cover & snap (x2)!!
Today New cable types Without braid Unshielded pairs Four-pair (not 2-pair) Research New connector types Die-cast (without a can ) New installation tools Simplify procedure Faster
Experiment - Subjects Shielded Termination IBM Type 1 2-pair S/FTP Category 6A 4-pair F/UTP Unshielded Termination Category 6 4-pair U/UTP Category 6A 4-pair U/UTP
Experiment - Setup Experienced Technicians Measure Time For Cable Prep Connector Prep Termination Dependent Variables Cabling Type Termination Tool
Termination Tools Punchdown Tool Each wire individually Installation Tool All wires simultaneously
Results Category 6 Unshielded Cable Type Tool Type Avg. Install Time Cat6 U/UTP Punchdown 2 minutes IBM Type 1 Pliers/Snips 3 min 45 sec (2pr) Two times myth plausible for IBM Type 1 cable termination Maybe even 4 times four pairs in over 7 minutes!
Results Category 6A Punchdown Cable Type Tool Type Avg. Install Time Cat6 U/UTP Punchdown 2 minutes Cable Type Tool Type Avg. Install Time Cat6A U/UTP Punchdown 3 min 15sec Cat6A F/UTP Punchdown 3 min 20sec Two times myth plausible for punchdown cable termination More of a Cat 6A issue than shielded issue Tighter twists, larger wires, more fillers
Results Category 6A Installation Tool Cable Type Tool Type Avg. Install Time Cat6 U/UTP Punchdown 2 minutes Cat6A U/UTP Punchdown 3 min 15 sec Cat6A U/UTP Install Tool 1 min 25 sec Cat6A F/UTP Install Tool 1 min 30 sec Even less than Category 6 punchdown!
Result: The Termination Time Myth BUSTED!!
Additional Considerations Field Testing could make installation time worse for unshielded depends on users choice of ANEXT testing options F/UTP vs. SF/FTP (Category 7) Category 7 (SF/FTP) would take longer Bonding and Grounding No additional time compared to unshielded Except for pulling back shield foil
The Antenna Effect
Myth The Antenna Effect The shield of a shielded cable acts like and antenna and collects radio frequency noise if it isn t grounded on both ends
Experiment - Subjects U/UTP Cable F/UTP Cable
Noise Source WAP (2GHz) Laptop PC Experiment - Setup U/UTP Cable F/UTP Cable Network Analyzer
Results Noise on Unshielded Cable
Results Noise on F/UTP Cable with Both Ends Grounded
Results Noise on F/UTP Cable with Single-ended Ground
Results Noise on Ungrounded F/UTP Cable
Result: The Antenna Effect BUSTED!!
Additional Considerations All metallic elements act as antennas to some extent even pairs in an unshielded cable The shield dramatically reduces the effects of RFI on the pairs, even if grounded on only one end or if ungrounded! But ground it anyway! A shield reduces the effects of induced noise for signals > 30 MHz Shield thickness property
Ground Loops
Myth Ground Loops Conductive loops transmit spurious current (noise) if there is any variation in voltage along the loop
Research Myth probably outgrowth of mains hum Two pieces of audio equipment plugged into different power outlets with different ground potentials This potential difference causes a spurious current to flow through the cables, creating an audible buzz
Experiment - Subjects F/UTP Cable
Experiment - Setup Apply 60 Hz AC current directly onto shield of F/UTP cable Not an induced current a placed current Test signal integrity on balanced twisted-pairs for presence of 60Hz influence
No influence on signal integrity at 60Hz Results
Result: Ground Loops BUSTED!!
Additional Considerations Basically proves that balance works, at least for 60Hz (low frequency) Audio cables tend not to be balanced or shielded Higher frequency effects reduced by the shield
The Baked Potato Effect
Myth The Baked Potato Effect The shield around a shielded cable is just like aluminum foil around a baked potato it traps generated heat and cooks the cable
Research Wrapping the potato in aluminum foil before cooking will help to retain moisture When cooking over an open fire or in coals, a jacket of foil prevents burning Aluminum foil is designed to keep moisture in the baked potato not to retain heat
Research Power over Ethernet Data cables used for signal and for power Power current creates heat via I 2 R losses Power over Ethernet Plus More power more current more heat
Experiment - Subjects U/UTP Cable F/UTP Cable
Cable bundles Experiment - Setup In insulated conduit, with firestopping
Worst case Experiment - Setup Place 720 ma current (original PoEP objective) The objective for PoEP is currently only 600mA 57 V dc measured at output end of cable (not source) Input Power on all four pairs Output Pairs wired in series
Experiment - Setup Thermocouple #3 Cable on Inner Layer of Bundle Thermocouple #2 Under Cable Jacket Thermocouple #4 Cable on Outer Layer of Bundle Thermocouple #1 Center of Cable Core
Results F/UTP Shielded Cable 7.0 6.0 5.0 4.0 3.0 2.0 1.0 0.0 0.25 0.75 1.25 1.75 2.25 2.75 3.25 3.75 Temperature Increase (C) Core Under Jacket Center Cable Outer Cable Time [hrs]
Results F/UTP and U/UTP Cable T-Rise measurements Ambient Temp. = 22.5 C T-Rise measurements Ambient Temp. = 21.7 C Center cable T-Rise Outer cable T-Rise Center cable T-Rise Outer cable T-Rise 10 10 Temperature Increase ( C) 8 6 4 2 Temperature Increase ( C) 8 6 4 2 0 0 Time Time F/UTP Cable U/UTP Cable
Results 720mA Cable Type Steady-state Temperature Rise F/UTP 6 C U/UTP 7 C Both show increase Shielded compares favorably
Result: The Baked Potato Effect BUSTED!!
Additional Considerations Unlikely to see that kind of temperature rise in the field Lower current than POEP objective Violation of fill ratios All four pairs carrying max power Max power measured at device end, not source end
Conclusions
Review Myth Conclusion Termination Time = x2 BUSTED!! Antenna Effects = Bad BUSTED!! Ground Loops = Bad BUSTED!! Baked Potato Effect = Hot BUSTED!!
Thank You! Remember: These are trained professionals do not try this at home! No animals (or potatoes) were harmed in this production